[ELF][PPC64] Add R_PPC64_DTPREL16* relocations for local-dynamic TLS model

[MaskRay]

Test comment

[MaskRay]

.

[MaskRay]

Also add R_PPC64_DTPREL16_LO_DS which is required to link llvm:

R_PPC64_DTPREL16_HA 77
R_PPC64_DTPREL16_LO_DS 102

[sfertile]

Hi @MaskRay,

we have already have a patch that adds support for nearly all the DTPREL relocations here: https://reviews.llvm.org/D47596. Its been sitting for a while with-out a review though, rather then submit a new patch that adds support for only a few of the local-dynamic relocations I would appreciate a review on that so we can commit support for all at once.

ELF/Arch/PPC64.cpp

uint32_t PPC64::calcEFlags() const {
   }
 
   return 2;
 }
 
 RelExpr PPC64::getRelExpr(RelType Type, const Symbol &S,
                           const uint8_t *Loc) const {
   switch (Type) {
+  case R_PPC64_DTPREL16:
+  case R_PPC64_DTPREL16_HA:
+  case R_PPC64_DTPREL16_LO:
+  case R_PPC64_DTPREL16_LO_DS:
+    return R_ABS;
   case R_PPC64_TOC16:
   case R_PPC64_TOC16_DS:
   case R_PPC64_TOC16_HA:
   case R_PPC64_TOC16_HI:
   case R_PPC64_TOC16_LO:
   case R_PPC64_TOC16_LO_DS:
     return R_GOTREL;
   case R_PPC64_TOC:
void PPC64::writePlt(uint8_t *Buf, uint64_t GotPltEntryAddr,
                      uint64_t PltEntryAddr, int32_t Index,
                      unsigned RelOff) const {
  int32_t Offset = PltHeaderSize + Index * PltEntrySize;
  // bl __glink_PLTresolve
  write32(Buf, 0x48000000 | ((-Offset) & 0x03FFFFFc));
 }
 
 static std::pair<RelType, uint64_t> toAddr16Rel(RelType Type, uint64_t Val) {
+  // For R_PPC64_DTPREL16* relocations, DTV pointer points 0x8000 bytes past the
+  // start of each TLS block. We also subtract 0x8000 (which equals to
+  // PPC64TocOffset) from Val to calculate the offset and thus can use the same
+  // formula.
   uint64_t V = Val - PPC64TocOffset;
   switch (Type) {
+  case R_PPC64_DTPREL16:
   case R_PPC64_GOT_TLSGD16:
   case R_PPC64_GOT_TLSLD16:
   case R_PPC64_TOC16:
     return {R_PPC64_ADDR16, V};
   case R_PPC64_TOC16_DS:
   case R_PPC64_GOT_TPREL16_DS:
     return {R_PPC64_ADDR16_DS, V};
+  case R_PPC64_DTPREL16_HA:
   case R_PPC64_GOT_TLSGD16_HA:
   case R_PPC64_GOT_TLSLD16_HA:
   case R_PPC64_GOT_TPREL16_HA:
   case R_PPC64_TOC16_HA:
     return {R_PPC64_ADDR16_HA, V};
+  case R_PPC64_DTPREL16_HI:
   case R_PPC64_GOT_TLSGD16_HI:
   case R_PPC64_GOT_TLSLD16_HI:
   case R_PPC64_GOT_TPREL16_HI:
   case R_PPC64_TOC16_HI:
     return {R_PPC64_ADDR16_HI, V};
+  case R_PPC64_DTPREL16_LO:
   case R_PPC64_GOT_TLSGD16_LO:
   case R_PPC64_GOT_TLSLD16_LO:
   case R_PPC64_TOC16_LO:
     return {R_PPC64_ADDR16_LO, V};
+  case R_PPC64_DTPREL16_LO_DS:
   case R_PPC64_TOC16_LO_DS:
   case R_PPC64_GOT_TPREL16_LO_DS:
     return {R_PPC64_ADDR16_LO_DS, V};
   default:
     return {Type, Val};
   }
 }
 

test/ELF/ppc64-local-dynamic.s

.Lfunc_lep0:
         .localentry     test, .Lfunc_lep0-.Lfunc_gep0
         mflr 0
         std 0, 16(1)
         stdu 1, -32(1)
         addis 3, 2, i@got@tlsld@ha
         addi 3, 3, i@got@tlsld@l
         bl __tls_get_addr(i@tlsld)
         nop
+        addi 4, 3, i@dtprel
+        addis 4, 3, j@dtprel@ha
+        addi 4, 4, j@dtprel@l
         ld 0, 16(1)
         mtlr 0
         blr
 
         .globl test_hi
         .p2align 4
         .type test_hi,@function
 test_hi:
k:
         .quad   66
         .size   k, 8
 
 // Verify that the input contains all the R_PPC64_GOT_TLSLD16* relocations.
 // InputRelocs: Relocation section '.rela.text'
 // InputRelocs:     R_PPC64_GOT_TLSLD16_HA 0000000000000000 i + 0
 // InputRelocs:     R_PPC64_GOT_TLSLD16_LO 0000000000000000 i + 0
 // InputRelocs:     R_PPC64_TLSLD          0000000000000000 i + 0
+// InputRelocs:     R_PPC64_DTPREL16       0000000000000000 i + 0
+// InputRelocs:     R_PPC64_DTPREL16_HA    0000000000000000 j + 0
+// InputRelocs:     R_PPC64_DTPREL16_LO    0000000000000000 j + 0
 // InputRelocs:     R_PPC64_GOT_TLSLD16_HI 0000000000000000 j + 0
 // InputRelocs:     R_PPC64_GOT_TLSLD16    0000000000000008 k + 0
 
 // The local dynamic version of tls needs to use the same mechanism to look up
 // a variables address as general-dynamic. ie a call to __tls_get_addr with the
 // address of a tls_index struct as the argument. However for local-dynamic
 // variables  all will have the same ti_module, and the offset field is left as
 // as 0, so the same struct can be used for every local-dynamic variable
 // used in the shared-object.
 // OutputRelocs:      Relocation section '.rela.dyn' at offset 0x{{[0-9a-f]+}} contains 1 entries:
 // OutputRelocs-NEXT: Offset Info Type Symbol's Value Symbol's Name + Addend
 // OutputRelocs-NEXT: R_PPC64_DTPMOD64
 
 // Check that the got has 3 entries, 1 for the TOC and 1 stucture of 2 entries
 // for the tls variables. Also verify the address so we can check the offsets
 // we calculate for each relocation type.
 // CheckGot: got          00000018 0000000000020100
 
 // got starts at 0x20100 so .TOC. will be 0x28100, and the tls_index struct is
 // at 0x20108.
 
 // #ha(i@got@tlsld) --> (0x20108 - 0x28100 + 0x8000) >> 16 = 0
 // #lo(i@got@tlsld) --> (0x20108 - 0x28100) = -7ff8 = -32760
+// i@dtprel --> (0x0 - 0x8000) = -32768
+// #ha(j@dtprel) --> (0x10 - 0x8000 + 0x8000) >> 16 = 0
+// #lo(j@dtprel) --> (0x10 - 0x8000) & 0xffff = -32752
 // Dis:     test:
 // Dis:        addis 3, 2, 0
 // Dis-NEXT:   addi 3, 3, -32760
+// Dis:        addi 4, 3, -32768
+// Dis-NEXT:   addis 4, 3, 0
+// Dis-NEXT:   addi 4, 4, -32752
 
 // #hi(j@got@tlsld) --> (0x20108 - 0x28100 ) > 16 = -1
 // Dis: test_hi:
 // Dis:   lis 3, -1
 
 // k@got@tlsld --> (0x20108 - 0x28100) = -7ff8 = -32760
 // Dis: test_16:
 // Dis:   li 3, -32760